Point-Cloud Wall Detection for Safe Transport-Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large transport vehicles face an increased risk of accidental contact with container and aisle walls due to their upsizing, especially when navigating areas with complex geometries such as corrugated panels or protrusions, as existing wall detection systems struggle to accurately model these surfaces.
Innovation Solution
A wall detection device equipped with a laser range finder or time-of-flight camera acquires a point cloud of wall coordinates, uses RANSAC for initial detection, and modifies wall positions to ensure parallel alignment, reducing the risk of contact by defining a safe movable range between walls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the transport vehicle is upsized to enhance cargo transport capacity, then the productivity is improved, but the risk of accidental contact with walls increases
Solution Approach 1:
The wall detection device performs preliminary detection of wall surfaces before the transport vehicle reaches them. The sensor detects wall surfaces ahead of time, and the controller pre-calculates safe movement ranges and adjusts the vehicle's path to avoid contact, allowing the upsized vehicle to navigate safely despite its larger dimensions.
2Device complexity
If the wall detection device uses simple geometric modeling, then the device complexity is reduced, but the measurement precision of wall surfaces with complex geometries deteriorates
Solution Approach 1:
Instead of using fixed geometric models, the system dynamically changes detection parameters by scanning multiple angles and positions. The sensor takes measurements at different angles and the controller processes these varying parameters to reconstruct the actual complex wall surface geometry, achieving high precision without requiring pre-defined geometric models.
Solution Approach 2:
The wall surface is divided into multiple detection points and segments rather than treating it as a single geometric entity. The sensor scans the wall at multiple positions and the controller processes each segment independently, allowing accurate representation of complex geometries like corrugated panels and protrusions without requiring a complete geometric model in advance.
3Measurement precision
If the sensor detects wall surfaces at multiple angles to capture complex geometries, then the measurement precision is improved, but the loss of time for data acquisition increases
Solution Approach 1:
The system performs detection at multiple angles and positions (excessive action) to ensure complete capture of complex wall geometries. The controller processes all collected data points to reconstruct the wall surface accurately, accepting the time cost as necessary for safe operation with upsized vehicles that have larger movement ranges.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The device accurately detects complex wall geometries, minimizing the risk of vehicle contact by calculating a safe path between walls, even with corrugated panels or protrusions, enhancing navigation accuracy and safety.
Implementation Method 1
a wall detection device having a sensor such as a laser range finder (LRF)
Implementation Method 2
A wall detection device equipped with a laser range finder or time-of-flight camera acquires a point cloud of wall coordinates
Data Source
AI summary
According to one embodiment, a wall detection device includes an acquirer and processing circuitry. The acquirer acquires a point cloud, which includes a series of coordinates of a plurality of points corresponding to a first wall and a second wall that oppose each other. The processing circuitry detects a first detected wall and a second detected wall based on a model and the acquired point cloud, the model representing a first plane which corresponds to the first detected wall and indicates a surface of the first wall and a second plane which corresponds to the second detected wall and indicates a surface of the second wall, and the model representing the first plane and the second plane being parallel to each other.